<html><head><meta http-equiv="Content-Type" content="text/html charset=us-ascii"></head><body style="word-wrap: break-word; -webkit-nbsp-mode: space; -webkit-line-break: after-white-space;" class=""><div class=""><div class=""><span class="" style="font-size: 14px;"><b class="" style="font-family: LucidaGrande; background-color: rgba(255, 255, 255, 0);"><u class="">Department of Computer Science </u></b><b class="" style="background-color: rgba(255, 255, 255, 0); font-family: LucidaGrande;"><u class="">Seminar</u></b></span></div></div><div class="" style="font-family: LucidaGrande;"><span class="" style="background-color: rgba(255, 255, 255, 0);"><br class=""></span></div><div class=""><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">Monday, April 17, 2017</span></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">2:30 pm</span></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">Ryerson 251 </span></div><div class="" style="font-family: LucidaGrande;"><br class=""></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">Kevin Jamieson</span></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">(UC Berkeley)</span></div><div class="" style="font-family: LucidaGrande;"><br class=""></div><div class=""><span class="" style="font-size: 14px;"><span class="" style="font-family: LucidaGrande;">Title: </span><span style="font-family: LucidaGrande;" class="">Efficient scalable algorithms for adaptive data collection</span></span></div><div class=""><span class="" style="font-size: 14px;"><font face="LucidaGrande" class=""><br class="gmail-m_4827095062818533991gmail_msg"></font><span class="" style="font-family: LucidaGrande;">Abstract: </span></span></div><div class=""><font face="LucidaGrande" class=""><span class=""><span style="font-size: 14px;" class="">In many applications, data-driven discovery is limited by the rate of data collection: the skilled labor it takes to operate a pipette, the time to execute a long-running physics simulation, the patience of an infant to remain still in an MRI, or the cost of labeling large corpuses of complex images. A powerful paradigm to extract the most information with such limited resources is active learning, or adaptive data collection, which leverages already-collected data to guide future measurements in a closed loop. But being convinced that data-collection should be adaptive is not the same thing as knowing how to adapt in a way that is both sample efficient and reliable. In this talk, I will present several examples of my provably reliable -- and practical -- adaptive data collection algorithms being applied in the real-world. In particular, I will show how my adaptive algorithms are used each week to crowd-source the winner of the New Yorker Magazine Cartoon Caption Contest. I will also discuss my application of adaptive learning concepts at Google to accelerate the tuning of deep networks in a highly parallelized environment of thousands of GPUs.<br class=""><br class=""></span><span style="font-size: 14px;" class="">-----</span></span></font></div><div class=""><font face="LucidaGrande" class=""><span class=""><span class="" style="font-size: 14px;"> </span><span class="" style="font-size: 14px;"><br class=""></span></span></font><span class="" style="font-size: 14px;"><span class="" style="font-family: 'Lucida Grande';">Bio: </span></span></div><div class=""><div class=""><div class=""><span style="font-family: LucidaGrande; font-size: 14px;" class="">Kevin Jamieson is a postdoctoral researcher working with Professor Benjamin Recht in the Department of Electrical Engineering and Computer Sciences at the University of California,</span><span style="font-family: LucidaGrande; font-size: 14px;" class=""> </span><span style="font-family: LucidaGrande; font-size: 14px;" class="">Berkeley. He is interested in the theory and practice of machine learning algorithms that sequentially collect data using an adaptive strategy. This includes active learning, multi-armed bandit</span><span style="font-family: LucidaGrande; font-size: 14px;" class=""> </span><span style="font-family: LucidaGrande; font-size: 14px;" class="">problems, and stochastic optimization. Kevin received his Ph.D. from the University of Wisconsin - Madison under the advisement of Robert Nowak. Prior to his doctoral work, Kevin received</span><span style="font-family: LucidaGrande; font-size: 14px;" class=""> </span><span style="font-family: LucidaGrande; font-size: 14px;" class="">his B.S. from the University of Washington, and an M.S. from Columbia University, both in electrical engineering. </span></div></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;"><br class=""></span></div><div class="" style="font-family: LucidaGrande;"><span class="" style="font-size: 14px;">Host: Risi Kondor</span></div></div></div></body></html>